1. The present study was designed to investigate the influence of long-term systemic treatment with Mycobacterium bovis bacillus Calmette-Guérin (BCG, 1 dose per animal, containing 6 x 10(4) colony-forming-units (CFu), 5 to 75 days beforehand) on oedema formation induced by intradermal injection of B1 and B2 selective agonists. The interaction between the B1 agonist des-Arg9-bradykinin and bradykinin was also investigated. 2. Intradermal injection (i.d.) of the B2 selective agonist tyrosine8-bradykinin (1-10 nmol) in naive (saline pretreated) animals, or in animals that had received BCG (30 days beforehand), caused dose-related and very similar oedema formation (ED50; 1.1 and 1.0 nmol/paw, respectively). I.d. injection of the selective B1 agonists des-Arg9-bradykinin (100 nmol) or des-Arg10-kallidin in naive animals caused very little paw oedema (0.04 +/- 0.06 and 0.07 +/- 0.02 ml, respectively, n = 5). However, i.d. injection of des-Arg9-bradykinin (10-300 nmol) or des-Arg10-kallidin (3-100 nmol) in animals pretreated with BCG, 30 days previously, resulted in dose-related and marked oedema formation, with mean ED50 values of 20.1 and 5.5 nmol/paw, respectively. 3. Oedema caused by i.d. injection of des-Arg9-bradykinin (100 nmol/paw) in rats pretreated with BCG was evident 5 days after treatment, reaching the maximum 30 days later, remaining stable for up to 45 days, and reduced markedly at 75 days. 4. The i.d. co-injection of the selective B1 antagonists des-Arg9[Leu8]-bradykinin (200 nmol), des-Arg10[Leu9]-bradykinin (30 nmol) and des-Arg9-NPC 17731 (30 nmol) significantly (18 +/- 3, 34 +/- 2 and 56 +/- 4%, respectively) prevented the paw oedema caused by i.d. injection of des-Arg9-bradykinin (100 nmol) in rats treated with BCG. These effects were selective, because the i.d. injection of the B1 selective antagonist des-Arg10[Leu9]-kallidin (30 nmol), at the same dose that consistently antagonized des-Arg9-bradykinin (100 nmol)-mediated paw oedema, had no significant effect against tyrosine8-bradykinin (3 nmol)-induced oedema in animals that had been treated previously with BCG. On the other hand, the i.d. co-injection of the selective B2 antagonist, Hoe 140 (10 nmol) at a dose which markedly inhibited tyrosine8-bradykinin (3 nmol)-induced oedema by 55 +/- 4%, did not significantly affect des-Arg9-bradykinin-induced paw oedema in animals pretreated with BCG. 5. Treatment of animals with dexamethasone (0.5 mg kg-1, s.c.) every 24 h, from day 0 to day 30, inhibited significantly (67 +/- 4%) the oedema caused by des-Arg9-bradykinin (100 nmol), but did not affect the paw oedema caused by tyrosine8-bradykinin (3 nmol) in animals pretreated with BCG. 6. Indomethacin (2 mg kg-1, i.p.), administered 1 h before experiments, significantly inhibited des-Arg9-bradykinin (100 nmol)-induced oedema formation, and, to a lesser extent, the paw oedema caused by tyrosine8-bradykinin (3 nmol) (44 +/- 4 and 20 +/- 4%, respectively). 7. These findings show that the long-term systemic treatment of rats with BCG promoted a time-dependent and consistent paw oedema formation to B1 agonists, des-Arg9-bradykinin and des-Arg10-kallidin, leaving responses to the B2 agonist tyrosine8-bradykinin unaffected. The upregulation of B1 receptors after BCG treatment was inhibited by dexamethasone, suggesting the possible involvement of de novo protein synthesis. Finally, our results also show that in BCG-treated animals, the B1 agonist des-Arg9-bradykinin interacts in a synergistic manner with bradykinin. Therefore, both B1 and B2 kinin receptors appear to play a relevant role in modulating chronic inflammatory processes.